OpenAI's o3 model has defeated Elon Musk's Grok 4 in a clean 4-0 sweep during the final of the first-ever AI chess tournament hosted by Kaggle. The event, held on Thursday, pitted general-purpose large language models against each other in a test of reasoning and strategy.
Both models advanced to the final after defeating competitors from Anthropic, DeepSeek, Google, and Moonshot AI. The matchup carried personal significance as Altman and Musk co-founded OpenAI a decade ago before Musk left to launch his own rival AI company, xAI. Their relationship has since soured, with Altman calling Musk a 'bully' after an unsuccessful takeover bid for OpenAI.
In response to Grok 4's performance, Musk downplayed the result, calling the AI model's ability a 'side effect'. The tournament marks the first major competition for large language models, which have become popular since ChatGPT's launch in 2022.
Commentating on the final, world chess champion Magnus Carlsen estimated Grok's rating at around 800, while OpenAI's model was around 1200. By comparison, Carlsen's peak rating is 2882. When asked on X, Grok itself estimated its rating at 1600-1800.
